For this assignment you must work in groups of three or four working with the Handyboard robot controller and the Legobug (or some variation) robot chassis. There are only three Handyboards, but enough parts to build 6 Legobugs. You must share the Handyboards.

Be careful with this equipment, it is expensive and delicate!!

This assignment has three main phases, due at different times. If folks get these going on time, we'll try some more difficult tasks, so don't procrastinate! The more difficult tasks are also more fun! At the end of this assignment, we'll have a public robot demonstration during and/or after class.

You must hand in one write-up per group for each phase. This write-up should describe your algorithms, your experience, and should include your (clean, commented, well-organized) code.

Unsolicited Advice: It is very easy to write hacked-up, "dirty" code when working with actual robots. You write a bit of code, then try it out, then tweak it. This will get you in a deep, deep hole when it comes to more sophisticated tasks. Use good programming style. Use data abstraction as you can. Good procedural abstraction will save you endless time! Comment your code, and keep around previous versions: don't overwrite unless you are sure you won't need it.

Phase I: Basic wandering

In this phase you will become familiar with the basics of Lego construction, and you will learn about Interactive C, the language for programming the Handyboard. You'll start to implement some very simple behaviors for the Robot.

• Build the "Legobug" according to the directions given separately. You will get started on this during class on Friday, January 30.
• Attach the Handyboard, and begin working with Interactive C. Try a few simple programs like Hello World. Try directly controlling the bug's motors interactively.
• Program the bug to move forward, then backward, then in a circle. Write a program to make the bug move in a square (on the table). Try that program with the bug on the floor.
• Make the bug wander randomly. Program it to stop when it hits an obstacle.
• Next, make the bug respond well to obstacles. It should back up, and turn to move in a new direction. Be sure it doesn't get stuck in corners, even when their angle is acute.
• The ultimate test of this phase is successful wandering in the bug "playpen" that I design. It will not be a simple square, nor will it be a tight maze, but it will include some weird angles.

Phase II: Follow the yellow-brick road

In this phase you will program your bug to use its light sensor(s) and follow a path marked on the ground. The bug will have to successfully stay on the path and get from a starting point to a goal. The path will be more constrained than in the previous section.

• First, design your light sensor configuration. How many light sensors will you use? Where will they be attached to the robot? Add the sensors to your bug.
• Next, program the robot to recognize the "path" from the "wilderness." Test it in a variety of lighting conditions.
• Program the bug to follow along straight sections of the path. Can it recognize when the path changes directions?
• Make the final product. The bug should be able to recover when the path changes direction, to find the path if it loses it. Try to ensure that it doesn't double back on the path it has already followed.

Phase III: Light-seeking

In this phase you will program your bug to navigate a playpen, seeking to approach a light source. When it has approached close to the light source, it should stop.

• Redesign your sensor configuration so that it will detect a light source that is close to surface level.
• Decide how to detect the direction of stronger light sources, and program the bug to move toward a light. You should be able to have the bug follow a flashlight around.
• Test your algorithms in the playpen, with a fixed light source and a complex walled environment.